24 Oct 2017

Some problems with peak water and decoupling

There are two definitions of "peak water," an idea borrowed from  "peak oil," which the fracking revolution -- as yet another example of profits from scarcity attracting innovation and thus abundance -- has effectively blown into pieces.

In the first, water demand exceeds renewable supplies, which requires that additional supply be "mined" from aquifers. This definition implies "unsustainable" because demand exceeds supply.

It's misleading because... read the post :)
In the second, used by the Pacific Institute and USGS, the term refers to a peak in demand for water in some or all sectors of the economy (see figure at right), which can be sustainable, as it refers to a "decoupling" of economic activity (e.g., GDP) from water use, a phenomenon also labeled as more crop per drop, doing more with less, and so on.

Back in 2009, Damian wrote how "peak water" is the wrong analogy to use with water because its physical and economic supply is much more the result of governance than physical or economic factors.*

These two definitions are simultaneously talking about the same thing and different things, which can be confusing. The first definition is more closely aligned with "supply," while the second is more about "demand." Although both can be used in any discussion of total water use, their conflation can lead to muddled thinking, e.g., falling groundwater levels (supply) leading to lower water use (demand), rather than lower water use resulting from a change in technology (demand) that ignores supplies.

Now, this short introduction puts the topic in context, but this post is really about the ridiculously bad data that are underlying the entire "debate." I came to this realization while trying to understand the "decoupling" cited by the Pacific Institute in a 2014 post, which shows a "break" around 1975-1980. What happened then? Did Reagan make some big change in economic policy that changed the nature of the economy or led to more sustainable water management?

After spending a few hours looking at data [here's my spreadsheet] and reports [past USGS reports], I am pretty sure that (1) peak water is not even well understood and (2) decoupling has long been the trend outside of agriculture but not within agriculture. Let's look at these in turn.

First, I tried to understand the 45 billion gallons/day (bgd) of "industrial use" from 1980 [see page 23 in the 1980 USGS report], but then I ran into two important issues with definitions. The first was that ALL these data are reporting "withdrawals" without regard to return flows. Even worse, "consumptive use" (the type that really does matter) is often calculated using ratios, leading to an estimated consumptive use of about 6 bgd. Second, the numbers combine freshwater & saline sources (39 & 6 bgd for withdrawals and 5 & 1 bdg for consumption, respectively), which also mixes metaphors in terms of sustainability (saline use is sustainable, by definition of "the ocean is really too big to drain"). These two examples should be enough to make you question any aggregated figures on peak demand, especially since REAL sustainability depends on local demand and supply.

Second, I used these figures (flawed as they are) to look into decoupling. I had a suspicion that much of the decoupling was related to the growth of services (tertiary production) at the expense of manufacturing (secondary) and agriculture (primary). Although I was right in a way, I was surprised to see such a strong trend in the data, i.e., that services and manufacturing have left agriculture in the dust mud in terms of value added to GDP (see same spreadsheet linked above for data) -- a fact that's obvious in this figure:

What does this figure show? That decoupling had occurred for manufacturing (averaging gains of 10 percent per five years, 1950-2010) and services (averaging 15 percent) but not for agriculture (averaging -1 percent). In other words, we're seeing "more crop per drop" for silicon chips but not for  potato chips.

Bottom Line: Do not trust aggregate measures of water use compared to accurate accounting at the watershed level. Do not mistake aggregate "decoupling" for improved (water) productivity in the agricultural sector. Farmers, as "residual claimants," are using as much water as they can get. Is that sustainable? I'm not sure, but missing and inaccurate data on withdrawals and consumptive use along with the entire absence of discussion of environmental water in the USGS report (the 2010 version!) tells me that we don't even know.

* Damian read this post and said: "Looks good to me. Those water figures from the USGS always were too aggregated to mean much to me, and given how unreliable CA water data is, and the lack of water data to begin with, it is surprising to base much on those numbers.

And then to the more general point about trying to find some peak - I agree as well. It's a muddled conversation, in part because of the data, but mostly because supply and demand matter, and trying to talk about one without the other is confusing."

No comments:

Post a Comment

Note: only a member of this blog may post a comment.